US8155763B2 - Operation control method, operation control device, and operation control system - Google Patents

Operation control method, operation control device, and operation control system Download PDF

Info

Publication number
US8155763B2
US8155763B2 US12/279,350 US27935007A US8155763B2 US 8155763 B2 US8155763 B2 US 8155763B2 US 27935007 A US27935007 A US 27935007A US 8155763 B2 US8155763 B2 US 8155763B2
Authority
US
United States
Prior art keywords
control
evaluation value
control object
model
deviation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related, expires
Application number
US12/279,350
Other languages
English (en)
Other versions
US20090012632A1 (en
Inventor
Takaaki Sekiai
Satoru Shimizu
Akihiro Yamada
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hitachi Ltd
Original Assignee
Hitachi Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hitachi Ltd filed Critical Hitachi Ltd
Assigned to HITACHI, LTD. reassignment HITACHI, LTD. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: SHIMIZU, SATORU, YAMADA, AKIHIRO, SEKIAI, TAKAAKI
Publication of US20090012632A1 publication Critical patent/US20090012632A1/en
Application granted granted Critical
Publication of US8155763B2 publication Critical patent/US8155763B2/en
Expired - Fee Related legal-status Critical Current
Adjusted expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
    • G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance

Definitions

  • the present invention relates to an operation control device/apparatus and method which adapt unsupervised learning.
  • the reinforcement learning is known as a framework of the learning control which generates an operation signal to environment such as a control object so that a measurement signal obtained from the environment may become desirable through a trial-and-error interaction with the environment.
  • the reinforcement learning has a learning function which generates an operation signal to the environment so that the expected value of the evaluation value obtained from the present state to the future may become the maximum, with a clue of a scalar evaluation value (in the reinforcement learning, called the reward) calculated using the measurement signal obtained from the environment.
  • Methods of implementing such a learning function include algorithms, such as Actor-Critic, Q-learning, and real-time Dynamic Programming, for example.
  • Dyna-architecture As a framework of the reinforcement learning into which the above-mentioned technique is developed. This is the method of learning beforehand what kind of operation signal should be better to be generated for a model which simulates a control object, and of determining the operation signal to be applied to the control object using this learning result. Dyna-architecture also has a model adjustment function which decreases an error between the control object and the model.
  • Patent Document 1 discloses the technology to which the reinforcement learning is applied.
  • two or more reinforcement learning modules which are a group of systems each possessing a model and a learning function.
  • a responsibility signal which takes a larger value for a smaller prediction error between the model and the control object in each of the reinforcement learning modules is calculated, and an operation signal generated from each of the reinforcement learning modules is weighted in proportion to the responsibility signal. In this way, the operation signal to be applied to the control object is determined.
  • Patent Document 1 JP-2000-35956A
  • the model which constitutes the above-mentioned Dyna-architecture might differ from the property of the control object.
  • the operation method might not become effective for the control object.
  • the operational condition of the control object might even get worse, if the operation signal based on the learned operation method is applied to the control object.
  • the present invention is made in view of the above circumstances and provides an operation control apparatus and an operation control method which allow operation of a control object without causing adverse affect on the operational condition of the control object, even when the deviation of the real system from the model (model error) arises.
  • An operation control method is employable in a control apparatus for controlling a control object by calculating operation amount to maximize or minimize an evaluation value obtained on the basis of a control deviation defined by deviation of control amount of the control object from a target value of the control object.
  • the operation control method includes the steps of: establishing a model for simulating a property of the control object; calculating operation amount to maximize or minimize an evaluation value on the basis of a control deviation of the model as a target; calculating an evaluation value based on a control deviation in controlling the control object by the operation amount; and determining operation amount change width defined by a difference between operation amount of a current step and operation amount to be determined at a next step, based on the deviation of the evaluation value of the control deviation of the control object from the evaluation value of the control deviation of the model.
  • control object can be operated without causing adverse affect on the operational condition of the control object, even when a model error arises.
  • FIG. 1 is a diagram illustrating a control apparatus according to an embodiment of the present invention applied to a control object
  • FIG. 2 is a table illustrating a mode of data stored in a generation parameter storing unit
  • FIG. 3 is a chart illustrating a screen image displayed on an image display apparatus
  • FIG. 4 is a flowchart illustrating processing of a generation parameter updating unit
  • FIG. 5 is a chart illustrating a property of a model
  • FIG. 6 is a chart illustrating an arrival point for every step
  • FIG. 7 is a chart illustrating difference between a property of the control object and a property of the model
  • FIG. 8 is a chart illustrating an operation method according to an embodiment of the present invention.
  • FIG. 9 is a chart illustrating a property of the model after modification
  • FIG. 10 is a chart illustrating an operation path after modification in controlling a control object
  • FIG. 11 is a chart illustrating relationship of a number of steps and an operation amount change width
  • FIG. 12 is a table illustrating an example of evaluation value calculation
  • FIG. 13 is a table illustrating an example of determining operation amount
  • FIG. 14 is a chart illustrating an example of a displayed screen of relationship of an operation amount space and an evaluation value.
  • 10 image display apparatus, 20 an external input apparatus, 30 : keyboard, 40 : mouse, 100 : control object, 200 : control apparatus, 300 : operation signal generator, 400 : model unit, 500 , 510 : evaluation value calculator, 600 : operation signal generation parameter storing unit, 700 : operation signal generation parameter updating unit, 800 : model parameter storing unit, 900 : model parameter updating unit
  • FIG. 1 is a diagram illustrating a control apparatus 200 according to an embodiment of the present invention applied to a control object 100 .
  • An operation signal generator 300 provided in the control apparatus 200 generates an operation signal 201 to be applied to the control object 100 .
  • An evaluation value calculator 500 calculates an evaluation value signal 203 using a measurement signal 202 obtained from the control object 100 .
  • the operation signal generator 300 receives the evaluation value signal 203 .
  • the operation signal generator 300 has a function for generating the operation signal 201 so that total of the expected value of the evaluation value signal 203 from the present state to the future may become the maximum or the minimum. The following explains the case where the operation signal generator 300 generates the operation signal 201 so that total of the expected value of the evaluation value signal 203 may become the maximum.
  • the evaluation value calculator 500 generates the evaluation value signal 203 corresponding to the deviation of the measurement signal 202 from the target value. For example, when the measurement signal 202 is in agreement with the target value, the evaluation value signal 203 is set to “1”, and when the measurement signal 202 is not in agreement with the target value, the evaluation value signal 203 is set to “0.” Alternatively, the evaluation value signal 203 is set such that the evaluation value signal 203 is in inverse proportion to the deviation of the measurement signal 202 from the target value. Namely, the evaluation value is closer to the target value as the numeric value is larger like +30, and the evaluation value is farther away from the target value as the numeric value is smaller like ⁇ 30, as described later in FIG. 5 .
  • the evaluation value calculation in this case can adopt plural methods.
  • An example of the evaluation value calculation is shown in FIG. 12 .
  • the example possesses a table in which the difference between the control amount and the target value and the evaluation value are listed in a corresponding manner.
  • the evaluation value can be generated with reference to the table.
  • the evaluation value can be calculated by setting the evaluation value as a function of the difference of the control amount and the target value.
  • the implementation of the operation signal generator 300 can be practiced by employing reinforcement learning.
  • the operation signal 201 is generated by trial and error in the early stage of learning. Then, the operation signal 201 is generated so that the evaluation value signal 203 may become larger as the learning is advanced.
  • Such a learning algorithm can employ algorithm such as Actor-Critic and Q-learning, for example.
  • the framework called Dyna-architecture is employed for the control apparatus shown in FIG. 1 .
  • the framework possesses a model unit 400 which simulates the control object 100 , and the operation signal generator 300 learns the generation method of the operation signal 201 for the model unit 400 beforehand, and generates the operation signal 201 using the result of learning.
  • the operation signal generator 300 possesses a function to generate an operation signal 204 to be fed to the model unit 400 , and to receive a measurement signal 205 and an evaluation value signal 206 from the model unit 400 .
  • the evaluation value signal 206 is calculated in an evaluation value calculator 510 using the measurement signal 205 .
  • the evaluation value calculator 510 possesses the same function as the evaluation value calculator 500 .
  • the operation signal generator 300 determines the operation signal 201 to be applied to the control object 100 , with reference to the data 207 stored in an operation signal generation parameter storing unit 600 .
  • FIG. 2 is a table illustrating a mode of data 610 stored in the operation signal generation parameter storing unit 600 .
  • the operation signal generation parameter storing unit 600 stores the data on the name of an operation terminal provided in the control object 100 , the operation amount change width per cycle, and the unit used.
  • the operation terminal can increase or decrease the operation amount in the range of the operation amount change width.
  • the number of the operation terminal may be one.
  • the operation amount change width is described for every operation terminal, alternatively, however, plural operation terminals can be put together to one group and the sum of the operation amount change width for the group of the operation terminals may be controlled within a limit.
  • the limiting value of the operation amount change width of FIG. 2 is determined in an operation signal generation parameter updating unit 700 .
  • a setting value necessary for the processing of the parameter update is inputted from an external input apparatus 20 possessing a keyboard 30 and a mouse 40 .
  • the information is displayed on an image display apparatus 10 such as CRT.
  • An operator of the control object 100 inputs a setting value 214 using the image display apparatus 10 and the external input apparatus 20 .
  • FIG. 3 is a chart illustrating a screen image displayed on an image display apparatus 10 .
  • the operator can set up a data 50 including an initial value, an upper limit, a lower limit, and an update ratio of the operation amount change width of the operation terminal.
  • the directions for use of the setting values set up here are explained using FIG. 4 .
  • FIG. 4 is a flowchart illustrating processing of the generating signal generation parameter updating unit 700 . In the following, the content of the processing in FIG. 4 is explained.
  • Processing 710 it is determined whether the number of steps t is greater than zero. When the number of steps is zero (0) (in the case of NO), Processing 720 is carried out, and when the number of steps is greater than zero (0) (in the case of YES), Processing 740 is carried out.
  • the number of steps is the number of times that the operation signal applied to the control object 100 is changed. The number of steps is zero (0) at the initial value, and increases by one (1) whenever an operation is practiced.
  • Processing 720 the initial value set up in FIG. 3 is acquired.
  • Processing 730 the initial value acquired in Processing 720 is sent to the generation parameter storing unit 600 as data 209 .
  • the last operation signal generation parameter stored in the generation parameter storing unit 600 is acquired as data 208 .
  • Equation 1 the operation amount change width is changed using Equation 1.
  • t stands for number of steps
  • G(t) stands for operation amount at step t
  • r 1 (t) stands for a value of the evaluation value signal 203
  • r 2 (t) stands for a value of the evaluation value signal 206
  • f(r 1 (t), r 2 (t)) is a function of variables r 1 (t) and r 2 (t).
  • G ( t+ 1) G ( t )+ f ( r 1 ( t ), r 2 ( t )) (Equation 1)
  • Equation 2 An example of the function f(r 1 (t), r 2 (t)) in Equation 1 is given by a function of Equation 2.
  • f ( r 1 ( t ), r 2 ( t )) ⁇ (
  • the operation amount change width may be calculated in the form of a function like Equation 2.
  • the operation amount may be determined with reference to a table stored, in which the difference of the evaluation value signals 203 and 206 and the operation amount change width G(t+1) ⁇ G(t) are tabulated in a corresponding manner as shown in FIG. 13 .
  • the control object can be operated without causing adverse affect on the operational condition of the control object. Furthermore, a flexible control is realizable according to deviation of the real system from the model.
  • the operation amount change width small when the difference of the evaluation values is large, and by making the operation amount change width large when the difference of the evaluation values is small, the operation amount can be changed safely when the deviation from the model is large, and the operation amount can be changed quickly when the deviation from the model is small.
  • a model parameter storing unit 800 the parameter necessary to constitute the model unit 400 is stored.
  • the model unit 400 is a physical model
  • the physical constants necessary to constitute the physical model are stored in the model parameter storing unit 800 .
  • the control object 100 is a thermal power generation plant
  • values such as a heat transfer rate are stored.
  • a model parameter updating unit 900 reads a parameter 212 stored in the model parameter storing unit 800 , modifies the parameter so that the properties of the control object and the model may be in agreement, sends a modified parameter 213 to the model parameter storing unit 800 , thereby updating the model parameter.
  • the model parameter updating unit 900 sets up a model parameter 211 to the model unit 400 , and updates the parameter of the model.
  • FIGS. 5-7 are charts explaining the problem expected to be generated when the control apparatus in related art is applied to the control object 100 .
  • FIG. 5 shows the relationship of the space of operation amount and the evaluation value obtained.
  • the evaluation value obtained is ⁇ 30 when the operation amount A 1 and B 1 are inputted into the model unit 400 .
  • the evaluation value obtained is +10 when the operation amount A 2 and B 2 is inputted.
  • Behavior that the total of the expected value of the evaluation value becomes the maximum avoids the area where the evaluation value is negative and goes to the area where the evaluation value is positive, taking a path such as indicated by the dotted line in FIG. 5 .
  • FIG. 6 is a chart illustrating an operation amount change width by one behavior shown by an arrow.
  • the operation amount change width in every behavior is set constant. In this way, the path from the starting point to the point where the evaluation value becomes +30 is reached by six steps.
  • FIG. 7 is a chart illustrating an example in which the model and the control object possess different properties.
  • the conditions of the operation amount that yields a negative evaluation value differ by the model and the control object.
  • the evaluation value after the first step becomes ⁇ 30, leading to an undesirable state.
  • FIGS. 8-10 are charts explaining the effect when the control apparatus according to an embodiment of the present invention is applied to a plant 100 .
  • the operation amount change width is not fixed but variably determined through the processing of FIG. 3 .
  • the operation amount change width of the first step can be made small.
  • the evaluation value after the first step becomes ⁇ 10.
  • the value ⁇ 10 is superior to the value ⁇ 30 which is obtained after the first step by the method in related art.
  • control object moves to the operational condition resembling the initial state; therefore, the safety of the control object can be maintained.
  • the control apparatus 200 obtains the information that the property of the control object 100 and the property of the model unit 400 are different.
  • the model parameter updating unit 900 updates the parameter stored in the model parameter storing unit 800 so that the property of the model unit 400 and the property of the control object 100 may be in agreement.
  • the operation signal 201 is returned so that the control object may return to the initial state (“Start” in FIG. 8 ).
  • control object can be controlled safely, by modifying the model when the difference of the evaluation values is greater than the predetermined value, or by following the model when the deviation of the real system from the model is smaller than the predetermined value.
  • FIG. 9 shows the relationship between the space of operation amount and the evaluation value obtained from the model after modification.
  • behavior that the total of the expected value of the evaluation value becomes the maximum avoids the area where the evaluation value is negative and goes to the area where the evaluation value is positive, taking a path such as indicated by the dotted line in FIG. 9 .
  • the operation path differs in the case where the model before modification is employed and in the case where the model after modification is employed.
  • FIG. 10 shows the path when controlling the control object 100 using the operation path after modification.
  • the operation amount change width is increased. As a result, as shown in FIG. 10 , the arrow becomes longer gradually.
  • FIG. 11 shows the relationship of the number of steps and the operation amount change width in the operation execution of FIG. 10 .
  • Equation 2 Since both the evaluation value of the model and the evaluation value from the control object are zero (0), the second term in Equation 2 is zero (0). Therefore, the operation amount change width increases by ⁇ per step.
  • the operation signal 201 is displayed on CRT 10 of FIG. 1 .
  • the data of the operation amount change width etc. which are the data 210 stored in the operating signal generation parameter storing unit 600 , can also be displayed.
  • the control amount 202 of the control object 100 can also be displayed.
  • CRT 10 can display on the screen the relationship of the space of operation amount and the evaluation value, as shown in FIGS. 5-10 .
  • FIG. 14 An example of the screen displaying the relationship of the space of operation amount and the evaluation value is shown in FIG. 14 .
  • the control apparatus 100 sets up the operation amount of plural operations applied to the control object on plural axes, respectively, and creates image information in which the start point and arrival point of each operation applied to the control object are displayed, and the arrival point of the operation at the previous step and the start point of the operation at the next step are connected.
  • the control apparatus 100 displays the created image information on CRT 10 . Accordingly, the amount of variation of each operation can be easily grasped by contrast with the whole operation.
  • the connection from the start point to the arrival point is displayed by an arrow.
  • the control apparatus 200 possesses the model 400 which simulates the property of the control object, the evaluation value calculator 510 of the model which calculates the evaluation value based on the control deviation in controlling the model as a target, and the evaluation value calculator 500 of the control object which calculates the evaluation value based on the control deviation in controlling the control object.
  • the control apparatus 200 calculates the difference of the evaluation value of the model and the evaluation value from the control object when each operation is performed, creates the display data to be displayed correspondingly to the display of each operation, and sends the created display data to CRT 10 .
  • the operation amount change width is made small immediately after the operation starts, and it is confirmed that the operation method learned for the model is effective also in the control object. Then, the operation amount change width is gradually increased, after it turns out that the property of the control object and the property of the model are analogous to each other and that the operation method learned for the model is effective also in the control object.
  • control object can be operated without causing adverse affect on the operational condition of the control object.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Feedback Control In General (AREA)
US12/279,350 2006-03-31 2007-01-18 Operation control method, operation control device, and operation control system Expired - Fee Related US8155763B2 (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
JP2006-096373 2006-03-31
JP2006096373A JP4952025B2 (ja) 2006-03-31 2006-03-31 運転制御方法，運転制御装置及び運転制御システム
PCT/JP2007/050682 WO2007116590A1 (ja) 2006-03-31 2007-01-18 運転制御方法，運転制御装置及び運転制御システム

Publications (2)

Publication Number Publication Date
US20090012632A1 US20090012632A1 (en) 2009-01-08
US8155763B2 true US8155763B2 (en) 2012-04-10

Family

ID=38580892

Family Applications (1)

Application Number Title Priority Date Filing Date
US12/279,350 Expired - Fee Related US8155763B2 (en) 2006-03-31 2007-01-18 Operation control method, operation control device, and operation control system

Country Status (4)

Country Link
US (1) US8155763B2 (zh)
JP (1) JP4952025B2 (zh)
CN (1) CN101390024A (zh)
WO (1) WO2007116590A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11173613B2 (en) * 2017-02-09 2021-11-16 Mitsubishi Electric Corporation Position control device and position control method

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4427074B2 (ja) 2007-06-07 2010-03-03 株式会社日立製作所 プラントの制御装置
JP4876057B2 (ja) * 2007-11-20 2012-02-15 株式会社日立製作所 プラントの制御装置、及び火力発電プラントの制御装置
US8135653B2 (en) 2007-11-20 2012-03-13 Hitachi, Ltd. Power plant control device which uses a model, a learning signal, a correction signal, and a manipulation signal
JP4627553B2 (ja) 2008-03-28 2011-02-09 株式会社日立製作所 プラントの制御装置および火力発電プラントの制御装置
CN105319966B (zh) * 2014-07-30 2017-10-20 南京南瑞继保电气有限公司 一种避免系统间通讯恢复后冗余系统同时退出值班的方法
JP6497367B2 (ja) * 2016-08-31 2019-04-10 横河電機株式会社 プラント制御装置、プラント制御方法、プラント制御プログラム及び記録媒体
JP6742222B2 (ja) * 2016-11-14 2020-08-19 株式会社日立製作所 運転支援装置及びプログラム
WO2018105412A1 (ja) * 2016-12-07 2018-06-14 ソニー株式会社 情報処理装置および方法、並びにプログラム
JP6662310B2 (ja) * 2017-01-11 2020-03-11 横河電機株式会社 データ処理装置、データ処理方法及びプログラム
CN111433694A (zh) * 2017-11-29 2020-07-17 三菱日立电力系统株式会社 运转条件评价装置、运转条件评价方法及发电设备的控制系统
JP6970078B2 (ja) 2018-11-28 2021-11-24 株式会社東芝 ロボット動作計画装置、ロボットシステム、および方法
JP7206874B2 (ja) * 2018-12-10 2023-01-18 富士電機株式会社 制御装置、制御方法及びプログラム

Citations (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH04256102A (ja) 1991-02-08 1992-09-10 Toshiba Corp モデル予測制御装置
US5166873A (en) * 1989-09-13 1992-11-24 Yokogawa Electric Corporation Process control device
JPH0695706A (ja) 1992-09-11 1994-04-08 Toshiba Corp モデル予測制御装置
US5347446A (en) 1991-02-08 1994-09-13 Kabushiki Kaisha Toshiba Model predictive control apparatus
US5448681A (en) * 1992-03-27 1995-09-05 National Semiconductor Corporation Intelligent controller with neural network and reinforcement learning
JPH0883104A (ja) 1994-09-12 1996-03-26 Toshiba Corp プラント制御装置
JPH09128011A (ja) 1995-10-31 1997-05-16 Fuji Electric Co Ltd モデル予測制御方法
JPH1074188A (ja) 1996-05-23 1998-03-17 Hitachi Ltd データ学習装置およびプラント制御装置
JPH10214112A (ja) 1997-01-29 1998-08-11 Hitachi Ltd 火力プラントシミュレータ
US5930136A (en) * 1990-06-04 1999-07-27 Hitachi, Ltd. Control device for controlling a controlled apparatus, and a control method therefor
JP2000035956A (ja) 1998-07-17 2000-02-02 Japan Science & Technology Corp エージェント学習装置
JP2001154705A (ja) 1999-11-25 2001-06-08 Hitachi Ltd 火力プラントの運転制御方法及び運転制御装置
JP2001236102A (ja) 2000-02-23 2001-08-31 Hitachi Ltd 制御装置および制御方法
JP2003005804A (ja) 2001-06-19 2003-01-08 Honda Motor Co Ltd プラントの制御装置
US20030120360A1 (en) 2001-04-20 2003-06-26 Yuji Yasui Plant control apparatus
JP2004094939A (ja) 2002-08-09 2004-03-25 Omron Corp モデル構造、制御装置、温度調節器および熱処理装置
US20060184465A1 (en) * 2004-10-21 2006-08-17 Hiroshi Tsujino Neural network element with reinforcement/attenuation learning
US20070022068A1 (en) * 2005-07-01 2007-01-25 Ralph Linsker Neural networks for prediction and control

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP3989775B2 (ja) * 2002-06-05 2007-10-10 株式会社日立製作所 制御モデルを用いた制御系のプリセット制御方法および制御装置
ITMI20021231A1 (it) * 2002-06-06 2003-12-09 Nuovo Pignone Spa Sistema di controllo e regolazione della temperatura di fiamma per turbine a gas monoalbero

Patent Citations (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5166873A (en) * 1989-09-13 1992-11-24 Yokogawa Electric Corporation Process control device
US5930136A (en) * 1990-06-04 1999-07-27 Hitachi, Ltd. Control device for controlling a controlled apparatus, and a control method therefor
JPH04256102A (ja) 1991-02-08 1992-09-10 Toshiba Corp モデル予測制御装置
US5347446A (en) 1991-02-08 1994-09-13 Kabushiki Kaisha Toshiba Model predictive control apparatus
US5448681A (en) * 1992-03-27 1995-09-05 National Semiconductor Corporation Intelligent controller with neural network and reinforcement learning
JPH0695706A (ja) 1992-09-11 1994-04-08 Toshiba Corp モデル予測制御装置
JPH0883104A (ja) 1994-09-12 1996-03-26 Toshiba Corp プラント制御装置
JPH09128011A (ja) 1995-10-31 1997-05-16 Fuji Electric Co Ltd モデル予測制御方法
JPH1074188A (ja) 1996-05-23 1998-03-17 Hitachi Ltd データ学習装置およびプラント制御装置
JPH10214112A (ja) 1997-01-29 1998-08-11 Hitachi Ltd 火力プラントシミュレータ
JP2000035956A (ja) 1998-07-17 2000-02-02 Japan Science & Technology Corp エージェント学習装置
US6529887B1 (en) 1998-07-17 2003-03-04 Agency Of Industrial Science And Technology Agent learning machine
JP2001154705A (ja) 1999-11-25 2001-06-08 Hitachi Ltd 火力プラントの運転制御方法及び運転制御装置
JP2001236102A (ja) 2000-02-23 2001-08-31 Hitachi Ltd 制御装置および制御方法
US20030120360A1 (en) 2001-04-20 2003-06-26 Yuji Yasui Plant control apparatus
US20060129250A1 (en) 2001-04-20 2006-06-15 Honda Giken Kogyo Kabushiki Kaisha Control system for plant
JP2003005804A (ja) 2001-06-19 2003-01-08 Honda Motor Co Ltd プラントの制御装置
JP2004094939A (ja) 2002-08-09 2004-03-25 Omron Corp モデル構造、制御装置、温度調節器および熱処理装置
US20060184465A1 (en) * 2004-10-21 2006-08-17 Hiroshi Tsujino Neural network element with reinforcement/attenuation learning
US20070022068A1 (en) * 2005-07-01 2007-01-25 Ralph Linsker Neural networks for prediction and control

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
Kyoka Gakushu, (Reinforcement Learning) published by Morikita Shuppan (Morikita Publishing) in Dec. 2000; authored by Richard S. Sutton and Andre G. Barto; translated jointly by Sadayoshi Mikami and Masaaki Minagawa.

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11173613B2 (en) * 2017-02-09 2021-11-16 Mitsubishi Electric Corporation Position control device and position control method

Also Published As

Publication number Publication date
JP4952025B2 (ja) 2012-06-13
US20090012632A1 (en) 2009-01-08
WO2007116590A1 (ja) 2007-10-18
CN101390024A (zh) 2009-03-18
JP2007272498A (ja) 2007-10-18

Similar Documents

Publication Publication Date Title
US8155763B2 (en) Operation control method, operation control device, and operation control system
KR101945863B1 (ko) 시뮬레이션 모듈을 사용하는 컴퓨터-지원 플랜트 제어 최적화를 위한 방법
CN108008627A (zh) 一种并行优化的强化学习自适应pid控制方法
WO2019098158A1 (ja) プラント管理システム及び管理装置
CN110989403B (zh) 综合能源调控系统及其控制方法、服务器
CN110018722B (zh) 用于热控制的机器学习装置、系统和方法
JP6946654B2 (ja) 制御装置、制御方法、および、制御プログラム
US11320811B2 (en) Plant operating condition setting support system, learning device, and operating condition setting support device
CN112859601B (zh) 机器人控制器设计方法、装置、设备及可读存储介质
KR20210013971A (ko) Pso를 이용한 모터 제어기의 자동 학습 튜닝 시스템
US20220326665A1 (en) Control system, and control method
JP2002367875A (ja) プロセス工程管理システムおよびプロセス工程管理方法
WO2021256135A1 (ja) 制御装置、方法及びプログラム
CN113655816B (zh) 钢包底吹氩系统流量控制方法及计算机可读存储介质
CN115562042A (zh) 一种分布式平行多层模型预测控制的发电控制方法
CN114880932A (zh) 一种电网运行环境模拟方法、系统、设备及介质
AU2021240175A1 (en) Model learning apparatus, control apparatus, model learning method and computer program
EP4050431A1 (en) Control device, method, program, and system
CN110778447A (zh) 一种混流式水轮发电机调速器空载pid参数优化方法
JP7462905B2 (ja) 制御装置、方法、プログラム及びシステム
US20240169209A1 (en) Computer-implemented method for producing a software image, suitable for a numerical simulation, of at least part of a real controller
CN118466173B (zh) 一种犬舍环境温度控制优化方法
CN118228781B (zh) 一种基于生成式对抗网络的训练系统
CN117193025B (zh) 加药机控制方法、装置、电子设备及存储介质
WO2023074163A1 (ja) 制御パラメータの生成方法、プログラム、記録媒体、および、制御パラメータの生成装置

Legal Events

Date Code Title Description
AS Assignment

Owner name: HITACHI, LTD., JAPAN

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:SEKIAI, TAKAAKI;SHIMIZU, SATORU;YAMADA, AKIHIRO;REEL/FRAME:021385/0910;SIGNING DATES FROM 20080723 TO 20080730

Owner name: HITACHI, LTD., JAPAN

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:SEKIAI, TAKAAKI;SHIMIZU, SATORU;YAMADA, AKIHIRO;SIGNING DATES FROM 20080723 TO 20080730;REEL/FRAME:021385/0910

FEPP Fee payment procedure

Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

ZAAA Notice of allowance and fees due

Free format text: ORIGINAL CODE: NOA

ZAAB Notice of allowance mailed

Free format text: ORIGINAL CODE: MN/=.

ZAAA Notice of allowance and fees due

Free format text: ORIGINAL CODE: NOA

STCF Information on status: patent grant

Free format text: PATENTED CASE

FPAY Fee payment

Year of fee payment: 4

MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 8TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1552); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

Year of fee payment: 8

FEPP Fee payment procedure

Free format text: MAINTENANCE FEE REMINDER MAILED (ORIGINAL EVENT CODE: REM.); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

LAPS Lapse for failure to pay maintenance fees

Free format text: PATENT EXPIRED FOR FAILURE TO PAY MAINTENANCE FEES (ORIGINAL EVENT CODE: EXP.); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY

STCH Information on status: patent discontinuation

Free format text: PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362

FP Lapsed due to failure to pay maintenance fee

Effective date: 20240410